Prices of wind power have been constantly decreasing of late, and now electricity from wind is almost as cheap as that from conventional fossil fuel power stations. It does however still suffer from the same problem of inconsistency. The wind just won’t play by the rules and blow 24 hours a day at a constant speed.
Turbines are designed to perform at an optimum wind speed. They then have complex control systems that can alter the magnetic torque of the turbine and tilt the angle of the blades in order to improve efficiency in low winds, or protect the turbine from damage during high winds. The systems tend to be complex and expensive, but a group of Chinese researchers have invented a new, cheap approach which could help bring the cost of wind turbines down even further, whilst actually increasing their efficiency.
The new system is based upon the human memory. Not in an AI sort of way, but by enabling the control system to memorise previous conditions and changes, so that they understand the optimum settings for each change in wind.
This new approach allows the control system to learn from old changes and adapt to new ones. It remembers old changes that were made in certain wind conditions, so that when those wind conditions are experienced again, it knows the most efficient settings for the turbine. In simulations the new system initially showed poor results, however with time it quickly learnt from its mistakes, improved its settings, and then matched traditional, expensive systems.
The researchers stated that, “the human-memory-based method holds great promise for enhancing the efficiency of wind power conversion.”
By. Joao Piexe of Oilprice.com
Joao is a writer for Oilprice.com